Beyond Reasoning Gains: Mitigating General Capabilities Forgetting in Large Reasoning Models

Open in new window